Marine Pollution Bulletin 62 (2011) 1822–1829

Contents lists available at ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier.com/locate/marpolbul

New threats of an old enemy: The distribution of the L. (: Teredinidae) related to climate change in the Port of area, the ⇑ Peter Paalvast a, , Gerard van der Velde b,c a Ecoconsult, Asterstraat 19, 3135 HA , The Netherlands b Radboud University Nijmegen, Institute for Water and Wetland Research, Department of Ecology and Ecophysiology, P.O. Box 9010, 6500 GL Nijmegen, The Netherlands c Netherlands Centre for Biodiversity Naturalis, P.O. Box 9517, 2300 RA Leiden, The Netherlands article info abstract

Keywords: The effects of four climate change scenarios for the Netherlands on the distribution of the shipworm Shipworm upstream of the – estuary are described. Global warming will cause dry and warmer sum- Salinity mers and decreased river discharges. This will extend the salinity gradient upstream in summer and fall River discharge and may lead to attacks on wooden structures by the shipworm. Scenarios including one or two degree Estuaries temperature increases by 2050 compared to 1990 with a weak change in the air circulation over Europe Global warming will lead to an increased chance of shipworm damage upstream from once in 36 years to once in 27 or Sea level rise 22 years, respectively; however, under a strong change in air circulation, the chance of shipworm damage increases to once in 6 or 3 years, respectively. The upstream expansion of the distribution of the ship- worm will also be manifested in other northwest European estuaries and will be even stronger in south- ern Europe. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction bodies in the Dutch Delta Area (Van Benthem Jutting, 1943; De Bruyne, 1994). (Teredinidae) are wood-boring marine bivalves. The If shipworms encounter optimal abiotic conditions, they are shipworm Teredo navalis L. appeared in 1730 in the coastal waters able to destroy fir piles 15 cm in diameter in six weeks (Snow, of the Netherlands and was described by Sellius (1733) as Teredo 1917), and even 10 m long, 25 cm thick oak pilings can be turned marina. Its origin is unknown, and the is therefore de- into rubble in 7 months (Cobb, 2002). In 1995, shipworm damage scribed as cryptogenic by Hoppe (2002). In the successive years in the US was estimated to cost approximately US$ 200 million of 1731 and 1732, massive destruction of the wooden construc- per year (Cohen and Carlton, 1995). Over the centuries, many at- tions that protected the dikes in Zealand and Westfrisland oc- tempts to protect wood structures from shipworm attacks have curred (Vrolik et al., 1860). Authorities attempted to use tropical more or less failed. They have used copper and lead plating, nails hardwoods and arsenic and to cover the wooden dike gates with with large flat heads (Teredo-nails), paraffin, tar, asphalt, and iron plates and nails, but the only (very expensive) solution was paints. The most effective deterrent, creosote, has been banned in changing the mode of dike construction. This began as soon as many countries because of its toxic and carcinogenic properties 1733 by defending the dikes with imported stones, and over the (Snow, 1917; Hoppe, 2002). Chemical impregnation with chrome centuries, this has led to the ‘‘petrification’’ of large parts of copper arsenate (CCA) or borax (CKB) is widely used as a wood pre- the Dutch coastline. Later outbreaks of the species took place in servative and is effective, although this practice remains controver- the Netherlands in 1770, 1827, 1858 and 1859. Not only were sial among environmentalists (Hoppe, 2002). sluices and dolphins in harbors upstream of the Rhine–Meuse The adult shipworm tolerates salinity conditions between 5 and estuary found to be infected in 1826, but quays also collapsed. 35 (Nair and Sarawathy, 1971) and thrives and reproduces at salin- Low river discharges resulting in increased salinity had created ities of 9 and higher (Kristensen, 1969; Soldatova, 1961 in Tuente favorable conditions for the shipworm. The shipworm is still com- et al., 2002). Boring activity stops below a salinity of 9–10. Pelagic monly found in Dutch coastal waters and enclosed salt water shipworm larvae survive at salinities as low as 6, and below that salinity level, pediveligers die within a few days (Hoagland, 1986). It was observed in Germany that a low salinity of 9 pre- ⇑ Corresponding author. Tel.: +31 107537549; fax: +31 847220307. vented the establishment of shipworm larvae (Hoppe, 2002). At E-mail addresses: [email protected] (P. Paalvast), g.vandervelde@ 1 À1 science.ru.nl (G. van der Velde). water current velocities exceeding 0.8 m s , shipworm larvae

0025-326X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2011.05.009 P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829 1823 can no longer attach to wood (Quayle, 1992). The optimal water ditions, with a Rhine river discharge at the Dutch–German border temperatures for growth and reproduction range between 15 and (Qbr) of 2200 m3 s À1, the complete salinity gradient from fresh to 25 °C. Spawning is initiated when the temperature rises above 11 seawater is found in this part of the estuary. This salinity gradient to 12 °C. These may breed from May until October (Grave, moves downstream at ebb tide and upstream at floodtide. The 1928). First year animals may reach sexual maturity in six weeks hydrology of the area is strongly controlled by means of the drain- (Lane, 1959). Temperatures up to 30 °C are tolerated. Boring activ- age program for the sluices (Fig. 1) based on the dis- ity decreases below 10 °C and stops at 5 °C(Roch, 1932; Norman, charge of the Rhine at the Dutch–German border (Qbr). The 1977). At temperatures near the freezing point, shipworms hiber- degree to which the sluice gates are opened at ebb tide increases nate until the water temperature becomes favorable again. with increasing discharge of fresh water. To avoid salinization up- The complete salinity gradient of the Rhine–Meuse estuary is stream of Rotterdam, 1500 m3 sÀ1 of fresh water is directed to the present in the area, and a vital shipworm popu- Port of Rotterdam area, i.e., the (1300 m3 sÀ1) lation exists in the large polyhaline harbors in its western region and the Hartelkanaal (200 m3 sÀ1). This flow of fresh water via (Beerkanaal and Calandkanaal). T. navalis settlement in the Port the Port of Rotterdam area can be maintained between a Qbr of of Rotterdam area decreases with increasing distance from the 1700 m3 sÀ1 and 4500 m3 sÀ1. Below 1700 m3 sÀ1, the salinity gra- sea floor in both test panels and beams (Paalvast and Van der Vel- dient shifts landward, while above 4500 m3 sÀ1 it moves seaward. de, 2011). This was also found by Scheltema and Truitt (1956) in At a Qbr of approximately 1100 m3 sÀ1, the Haringvliet sluices are test panels in coastal waters of Maryland in the US and by Tuente closed completely, and both Meuse and Rhine waters flow into the et al. (2002) in the harbors of Bremerhaven in Germany. sea at the Hook of Holland. The isohalines of the bottom water in The situation in the Rotterdam port area, which is one of the the area under average and low river discharges (1000 m3 sÀ1) largest ports in the world, can be used as a typical example of were calculated using the RIJMAMO model of Rijkswaterstaat the estuaries of all large temperate rivers in western Europe. Var- (Bol and Kraak, 1998)(Fig. 2A and B). ious climate models (Van den Hurk et al., 2006, 2007; Boé et al., The Rhine–Meuse estuary is microtidal, with an average tidal 2009; Diaz-Nieto and Wilby, 2005) used to predict the effects of range of 1.75 m at the Hook of Holland, which gradually decreases global warming on the hydrology of these large rivers show, on upstream to 1.10 m at Hagestein on the and 0.25 m at Heusden average, a serious decrease in precipitation during summer and on the Meuse. At springtide, the tidal cycle at the Hook of Holland fall, leading to lower river discharges. Combined with an expected exhibits approximately a 4 h flood period, a 4 h ebb period and a sea level rise, this will lead to increasing salinity over large parts of 4.5 h low-water period. the estuarine gradient (Beijk, 2008). Under conditions of average river discharge, the water above Therefore, the following research question was formulated: the sea floor of the harbors of the (Fig. 2A) in the what is the risk of shipworm damage (i.e., the expansion of its dis- eastern region is fresh to oligohaline, but at discharges below tribution upstream) under present climate conditions and under 1000 m3 sÀ1, the area becomes a-mesohaline, with salinities over climate change due to global warming? 10. When the discharge of the Rhine river remains below 1000 m3 sÀ1 for a few weeks during the shipworm breeding season, conditions might become favorable for the bivalves (Paalvast and 2. Materials and methods Van der Velde, 2011). Larvae transported with the tidal currents to the eastern part of the Port of Rotterdam area could settle and 2.1. Study area grow in the wooden oak and pine poles on which several old quays are build, particularly in harbors where the current velocity is con- The Port of Rotterdam is situated in the estuary of the Rhine and siderably reduced. Salinity depth profiles show a sharp increase in Meuse rivers (Fig. 1). It stretches over a length of 40 km and covers salinity one to two meters above the sea floor under low river dis- 10,500 hectares, 3440 hectares of which are covered by harbor charges (< 1000 m3 sÀ1), which makes this water layer suitable for waters and 1960 hectares, by rivers and canals. Under average con- the settlement and growth of the shipworm. In the summer of

Fig. 1. The Port of Rotterdam in the Rhine–Meuse Estuary. 1824 P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829

Fig. 2. (A) The 5 and 10 isohalines during ebb and flood periods in the Port of Rotterdam area under an average discharge of the Rhine at the Dutch–German border (Qbr) of 2200 m3 sÀ1. (B) The 5 and 10 isohalines during ebb and flood periods in the Port of Rotterdam area under a discharge of the Rhine at the Dutch–German border (Qbr) of 1000 m3 sÀ1.

2003, the discharge of the Rhine river dropped below 1000 m3 sÀ1, reaches a permanent level (9 or higher) corresponding to necessary and the salinity in the eastern part of the Port of Rotterdam area conditions for the shipworm to settle, and its growth was com- rose to levels above 10 (Paalvast and Van der Velde, 2011). puted (Fig. 3). Van den Hurk et al. (2006) of the Royal Netherlands Meteoro- 2.2. Risk analysis logical Institute have elaborated climate change scenarios for the Netherlands known as the KNMI’06 scenarios (Fig. 4). Two antici- SIMONA is the hydraulics information system of the Ministry of pated circulation regime changes are included in the scenarios: a Transport (Ministry of Transport, Public Works and Water Manage- strong circulation change, which induces warmer and moister win- ment, 2008), consisting of a collection of mathematical simulation ter seasons and increases the likelihood of dry, warm summer con- models that describes hydrodynamic processes. It includes four ditions, and a weak circulation change. Both regimes are presented models for the simulation of hydrodynamic phenomena, such as for +1 °C and +2 °C global temperature increases for 2050 com- simulations of tides, the transport of solutes in water, and water pared to 1990, producing a total of four scenarios. In the G scenar- movements, and it is designed as a layered system that makes ios, the maximum mean sea level rise is 25 cm, and in the W use of uniform data storage. The Port of Rotterdam Authority uses scenarios it is 35 cm. SIMONA for forecasting hydrodynamics 36 h in advance twice a Rhineflow (Van Deursen, 1995, 2003) is a spatially distributed day. water balance model of the Rhine basin that can simulate river Using the SIMONA modeling system with the (critical) river dis- flows, soil moisture, snow pack and groundwater storage with a charge (crit-Qbr) under the prevailing climate conditions in the 10 days time step. In order of the Ministry of Transport, Carthago eastern part of the Port of Rotterdam as far upstream as the Rijnha- Consultancy, Rotterdam, simulated the effects of these climate sce- ven (Fig. 2A), the salinity at one to two meters above the sea floor narios using this model in terms of the percentage change (Fig. 5) P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829 1825

Fig. 3. Example of the output of the SIMONA hydraulics information system at a Qbr of 700 m3 sÀ1 for the salinity at the sea floor of the Waalhaven in the eastern part of the Port of Rotterdam area.

in the daily river discharge (Van Deursen, 2006). This relative change in the discharge under each scenario was applied to mea- surements of the average daily discharge at the Dutch–German border (Qbr) over the period 1901–2008 to simulate the discharge under these scenarios. The crit-Qbr for the KNMI’06 scenarios in relation to rising sea level was elaborated. A period of 14 days of crit-Qbr was considered as the minimum for shipworm settlement and boring to cause damage, and a period of 42 days was considered for the worms to complete their lifecy- cle (Richards et al., 1984). By counting the number of days per year on which the dis- charge over the period 1901–2008 and the simulated discharges Fig. 4. Schematic overview of the four KNMI’06 scenarios (after Van den Hurk, et al., over the same period for each of the KNMI’06 scenarios reached 2006). the crit-Qbr or lower, the number of periods of 14 days or longer

Fig. 5. Simulated relative change in the discharge of the Rhine river as a consequence of the KNMI’06 scenarios (after Van Deursen, 2006). 1826 P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829 and their repetition time (risk of damage) were calculated. The Table 1 length of the periods was used to determine whether the ship- Number of years with periods of 14 days or longer with a crit-Qbr or lower and years with a Qbr < 600 m3 sÀ1 with a risk of shipworm damage under the present climate worm could complete its lifecycle in the eastern part of the Port conditions (1990) and the KNMI’06 scenarios for 2050, the length of the periods, their of Rotterdam under the KNMI’06 scenarios. repetition time and the number of periods > 42 days for the shipworm to complete its life cycle.

3. Results Climate scenarios Present GG+ WW+ Number of years 3 5 18 4 45 3.1. Risk of damage under the present conditions Number of extreme years Qbr 003011 < 600 m3 sÀ1 Using the SIMONA modeling system, it was calculated that for Number of periods 3 7 20 4 46 3 À1 Average length periods in days 21.7 25.1 53.7 52.0 57.0 the Port of Rotterdam area, a discharge of 700 m s or less Standard deviation 10.7 11.6 34.5 31.3 33.6 (crit-Qbr) of the Rhine river at the Dutch–German border under Maximum length in days 34.0 49.0 132.0 93.0 142.0 the prevailing climate conditions would lead, within a week, to Minimum length in days 15.0 16.0 14.0 27.0 14.0 salinities of the bottom water in the harbors of the Nieuwe Maas Median length in days 16.0 19.0 46.0 44.0 49.5 Repetition time in years 36 21.6 6 27 2.4 in the eastern part of the Port of Rotterdam upstream to the Rijn- Number of periods > 42 days 0 1 10 2 26 haven (Fig. 2A) appropriate for settlement and growth of the ship- worm (and, thus, expansion of its distribution upstream). From 1901 to 2008, the crit-Qbr was reached for a period of 2 weeks that creates conditions for the shipworm to settle and grow. or longer in only 3 years (Table 1, Fig. 6), which indicates a repeti- Using test panels placed on the sea floor, Paalvast and Van der tion time of 36 years, or that the risk of damage is once in 36 years. Velde (2011) demonstrated that with the flood stream, shipworm The longest uninterrupted period lasted 34 days, which is too short larvae could travel 20 km upstream to 0.5 km before the conflu- for the shipworm to complete its life cycle. ence of the Nieuwe Maas and during a period in which the discharge of the Rhine was slightly higher than 1000 m3 sÀ1. 3.2. Risk of damage due to climate change Under the G and W scenarios, the probability of shipworm attacks in the eastern part of the Port of Rotterdam increases slightly With the sea level rise under the KNMI’06 scenarios for 2050, compared with the present situation, but the periods with critical the required salinity conditions for the settlement and growth of discharges are becoming much longer, so the eventual damage to the shipworm in the eastern part of the Port of Rotterdam up- wooden constructions will also increase. stream to the Rijnhaven (and, thus, expansion of its distribution Under the G+ and W+ scenarios, the probability of the risk of upstream) will occur at a discharge of approximately 800 m3 sÀ1 shipworm attacks and damage to underwater fir and oak harbor or lower (crit-Qbr). structures in the eastern part of the Port of Rotterdam greatly in- Under the G and W scenarios, with a weak change in air circu- creases, not only due to a decrease in the repetition time, but also lation patterns, there is only a slight decrease in river discharge because the periods with a crit-Qbr or lower persist over 2 months in summer and fall (Fig. 5), and thus, the increase in the number on average, which is long enough for the shipworm to reach matu- of days with a crit-Qbr is small compared with the number of days rity and to reproduce. with a crit-Qbr observed over the period 1901–2008 (Figs. 6–8). Once settled, these animals are able to continue to grow and For the G and W scenarios, the crit-Qbr is reached for 14 days or thus cause damage, even with a discharge exceeding the crit-Qbr longer in only 5 and 4 years, respectively, resulting in risks of dam- 3 1 by 200 m sÀ . When the salinity drops below 9 during the ebb age once in 22 and 27 years (Table 1). Under both scenarios, the phase, shipworms close their boring tubes using their pallets and shipworm is able to complete its lifecycle in 1 or 2 years over a per- reopen them when the salinity conditions become favorable again iod of 108 years for G and W, respectively. under flood conditions. If the salinity during both ebb and flood Under a future climate with a strong change in air circulation periods remains too low, the animals will survive for 6 weeks be- patterns, scenario G+ leads to many days and to several periods fore they die. of 14 days or longer associated with a crit-Qbr in summer and fall Under the G+ and W+ scenarios, there are years presenting ex- (Table 1, Figs. 7 and 8) and a considerable risk of damage once in treme salinization of the eastern part of the Port of Rotterdam 6 years. In 50% of these periods, the shipworm could complete its 3 1 due to long periods with a Qbr of approximately 600 m sÀ , due lifecycle between one to three times in a year. In three periods, to which the oldest harbors of Rotterdam, approximately 1 km up- the crit-Qbr reaches values less than 600 m3 sÀ1 for several days. stream from the Rijnhaven, become at risk, which may possibly Under a future climate with a strong change in air circulation also affect the historic ships at the quays. patterns, scenario W+ will result in a period of 14 days or longer The shipworm will not be able to settle permanently in the east- with a crit-Qbr almost once in approximately 2 years, and once ern part of the Port of Rotterdam because of the large winter and in 4 years, the shipworm could complete its life cycle from one spring discharges that occur at present, and even greater dis- to three times in a year. charges due to climate changes will wash out the brackish water Under the G+ and W+ scenarios, there is no significant differ- from the harbors for a few months, which is a period that is much ence in the lengths of periods of potential shipworm damage too long for the shipworm to survive. (t(64) = 0.37, p > 0.7). À In addition to the eastern part of the Port of Rotterdam, in the shallower old harbors of Maassluis and Vlaardingen, the conditions 4. Discussion might become favorable under the G+ and W+ scenarios for the shipworm to settle and grow in the fir piles on which the old quays The increased risk of the extension of the shipworm towards the are built, and if they are not replaced by that time, the mechanical eastern part of the Port of Rotterdam as far as the Rijnhaven de- stability of the piles and, consequently, that of the quays will be pends greatly on a temperature increase with changed air circula- impaired. tion patterns. However, under the present situation, at discharges Paalvast and Van der Velde (2011) found an average body below 1000 m3 sÀ1, there is salinization above the sea floor in the length growth rate of 1.5 mm dayÀ1 in 2006 and wood consump- most downstream part of the eastern portion of the Port of Rotter- tion of 12.4% of fir wood panels in the polyhaline harbors in the P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829 1827

Fig. 6. Periods of discharges of the Rhine river at the Dutch–German border of 6 700 m3 sÀ1 during the period 1901–2008 and their occurrence in the breeding and growing season of the shipworm.

Fig. 7. Simulated periods of discharges of the Rhine river at the Dutch–German border of 6 800 m3 sÀ1 during the period 1901–2008 and their occurrence in the breeding and growing season of the shipworm as a consequence of the KNMI’06 scenarios.

western part of the Port of Rotterdam for first year shipworm indi- on wood after one month of exposure. These figures clearly show viduals in 130 days. The number of shipworm attacks per m2 was the potential threat of this species, and therefore, it is not surpris- approximately 250. Hoppe (2002) recorded 40,000 larvae per m2 ing that when shipworms encounter structures made of digestible 1828 P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829

This study on the effects of global warming due to climate change on the expansion of the distribution of the shipworm up- stream in the Rhine–Meuse estuary clearly shows how ecologists can benefit from computer models developed by meteorologists and hydrologists. Despite the limitations and uncertainties that each model has, they are invaluable in providing insight into the ecological impacts of climate change.

5. Conclusion

Climate change with a global increase in temperature of one or two degrees coinciding with a strong change in air circulation that leads to low river discharges in summer and fall will extend salin- ity gradients upstream in western Europe estuaries and consider- ably increase the risk of damage to wooden structures by the shipworm in the near future. Fig. 8. Cumulative sum of days with potential shipworm damage in the eastern part of the Port of Rotterdam under the prevailing climate conditions (C) and the Acknowledgements KNMI’06 scenarios as simulated for the period 1901–2008. wood into which they can bore, these are ruined in a short period We thank the Rotterdam Port Authority for financing the use of of time. The complete destruction of a US Navy base in the San the SIMONA modeling system. Francisco Bay over just two years from 1919–1920 serves as a good example of the destructive capabilities of these bivalves (Cohen, References 2004). However, not only climate change, but also the improve- ment of water quality will have an effect on the distribution of Admiraal, W., van der Velde, G., Smit, H., Cazemier, W.G., 1992. The rivers Rhine and Meuse in The Netherlands: present state and signs of ecological recovery. the shipworm in estuaries. After the Clean Water Act of 1972, pil- Hydrobiologia 265, 97–128. ings that stood in the Hudson estuary in New York, USA for Beijk, V., 2008. Klimaatsverandering en verzilting Modelstudie naar de effecten van 100 years began to fall, and more recently, parts of a wooden de KNMI’06 klimaatscenario’s op de verzilting van het noordelijk deltabekken. Rijkswaterstaat Waterdienst Rijkswaterstaat Dienst Zuid-Holland. Brooklyn pier collapsed due to shipworm damage. The city of Rapportnummer: 2008.035. New York has already spent hundreds of millions of dollars repair- Bij de Vaate, A., Breukel, R., van der Velde, G., 2006. Long-term developments in ing wooden support pilings (Cobb, 2002). Additionally, in the Rot- ecological rehabilitation of the main in the Rhine delta: fish and macroinvertebrates. Hydrobiologia 565, 229–242. terdam port area, the water and sediment quality has been Boé, J., Terray, L., Martin, E., Habets, F., 2009. Projected changes in components of improved since the 1970’s, when pollution was severe. This was the hydrological cycle in French river basins during the 21st century. Water due to decreasing the levels of pollutants in the Rhine river as a re- Resources Research 45. Article no. W08420. Bol, R., Kraak, A., 1998. MER Beheer Haringvlietsluizen. Over de grens van zout en sult of the Rhine Action Plan following the Sandoz accident in 1986 zoet. Deelrapport Water- en Zoutbeweging. ISBN: 903694871. RWS, (Admiraal et al., 1993; Bij de Vaate et al., 2006). notanummer: apv 98/093. 212 pp. Various types of tropical timber (hardwood) were tested over a Cobb, K., 2002. Return of a castaway the gripping story of a boring clam – shipworm. Science, News 162, 72. period of 30 years in the marine waters of the Netherlands (Kon- Cohen, A.N., 2004. Invasions in the sea. Park Science 22, 37–41. inklijk Instituut voor de Tropen, 1972). The results showed that Cohen, A.N, Carlton, J.T., 1995. Nonindigenous aquatic species in a United States not only the hardness of timber, but also the presence of silica par- estuary: a case study of the biological invasions of the San Francisco Bay and ticles and poisonous alkaloids in wood affords considerable protec- delta. NOAA: USA. 251 pp. De Bruyne, R.H., 1994. Schelpen van de Nederlandse kust. Stichting tion against shipworm attacks. However, even the most resistant Jeugdbondsuitgeverij, Utrecht, 2e druk, 165 pp. hardwoods showed at least some damage by the shipworm after Diaz-Nieto, J., Wilby, R.I., 2005. A comparison of statistical downscaling and climate 30 years of exposure. Additionally, the shipworm lives in the outer change factor methods: impacts on lowflows in the river Thames United Kingdom. Climatic Change 69, 245–268. 3 cm of hundreds of pilings, fender structures and piers made from Grave, B.H., 1942. The sexual cycle of the shipworm, Teredo navalis. Biol. Bull. 82, driftwood of the hardwood basralokus (Dicorynia paraensis Benth.) 438–445. in the polyhaline harbors of the Port of Rotterdam (Paalvast and Hoagland, K.E., 1986. Effects of temperature, salinity, and substratum on larvae of the shipworms Teredo bartschi Clapp and Teredo navalis Linnaeus (Bivalvia: Van der Velde, 2011). With a reproductive rate of up to 2 million Teredinidae). Am. Malacol.Bull. 4, 89–99. larvae per individual (Hoppe, 2002) and the ability to travel over Hoppe, K., 2002. Teredo navalis – The cryptogenic shipworm. In: Leppäkoski, E., large distances through the flood stream, this species is on standby Gollasch, S., Olenin, S. (Eds.), Invasive aquatic species of Europe, Distribution impacts and management. Kluwer, Academic Publishers, The Netherlands. to attack wooden structures upstream once the river discharge de- Koninklijk Instituut voor de Tropen, 1972. Shipworm experiments during the period creases and salinity rises. 1938-1970. A comparative study of the resistance of tropical timber to Furthermore, the increased length of the salinity gradient and, shipworm. Communication N0. 62 of the Department of Agricultural Reseach. Kristensen, E., 1969. Attacks by Teredo navalis L in inner Danish waters in relation to consequently, the extension of the distribution of the shipworm up- environmental factors. Vidensk. Meddr. Dansk naturh. Foren. 132, 199–210. stream that are predicted within the Rhine–Meuse estuary as a con- Lane, C.E., 1959. Some aspects of the general biology of Teredo, In: Ray, D.L. (Ed.) sequence of climate change due to global warming will be similarly (1959). Marine boring and fouling organisms: Symposium. Friday Harbor manifested in other northwest European estuaries, such as the Elbe, Symposia, 137–158. Ministry of Transport, Public Works and Water Management, 2008. WAQUA/ Thames and Loire. The impact of climate change in southern Euro- TRIWAQ - two- and three-dimensional shallow water flow model. Technical pean estuaries will be even stronger under the worst case scenario, documentation, version 3.3, March 2008. W+, as in summer and fall, the decrease in precipitation may reach as Nair, N.B., Saraswathy, M., 1971. The biology of wood-boring teredinid molluscs. Adv. Mar. Biol. 9, 335–509. much as 80% (Van den Hurk et al., 2006). When salinity drops to lev- Norman, E., 1977. The geographical distribution and the growth of the wood-boring els favorable for the shipworm to settle and grow in places where molluscs Teredo navalis L and megotara (Hanley) and Xylophaga digestible wooden constructions are located in southern estuaries, dorsalis (Turton) on the Swedish west coast. Ophelia 16, 233–250. Paalvast, P., van der Velde, G., 2011. Distribution, settlement and growth of first these constructions will eventually be attacked by the animal and year individuals of the shipworm Teredo navalis L (Bivalvia: Teredinidae) in the will be seriously impaired or may collapse. Port of Rotterdam area, the Netherlands. Int. Biodet. Biodegr. 65, 379–388. P. Paalvast, G. van der Velde / Marine Pollution Bulletin 62 (2011) 1822–1829 1829

Quayle, D.B., 1992. Marine wood borers in British Columbia. Canadian Special Hazeleger, W., Drijfhout, S., 2006. KNMI Climate Change Scenarios 2006 for the Publication of Fisheries and Aquatic Sciences 115, 1–55. Netherlands. KNMI Scientific Report WR 2006-01. Richards, B.R., Hillman, R.E., Maciolek, N.J., 1984. Shipworms. In: Kennish, M.J., Lutz, Van den Hurk, B., Klein Tank, A., Lenderink, G., van Ulden, A., van Oldenborgh, G.J., R.A. (Eds.). Lecture Notes on Coastal and Estuarine Studies - Ecology of Barnegat Katsman, C., van den Brink, H., Keller, F., Bessembinder, J., Burgers, G., Komen, Bay, New Jersey, Springer-Verlag, New York. pp. 201–225, 396 p. G., Hazeleger, W., Drijfhout, S., 2007. New climate change scenarios for the Roch, F., 1932. Einige Beobachtungen zur Ökologie und Physiologie von Teredo Netherlands. Water Science & Technology 56, 27–33. navalis L. Ark. Zool. 24, 1–17. Van Deursen, W.P.A., 1995. Geographical Information systems and dynamic models. Scheltema, R.S., Truitt, R.V., 1956. The shipworm Teredo navalis in Maryland coastal Development and application of a prototype spatial modeling language. PhD- waters. Ecology 37, 841–843. thesis, May 19, 1995. Faculty of Spatial Sciences University of Utrecht, The Sellius, G., 1733. Historia naturalis teridinis seu Xylophagi marini, tubule- Netherlands. conchoideis speciatim belgici cum tabulis ad vivum delineatis. Trajecti ad Van Deursen, W.P.A., 2003. Klimaatveranderingen in de stroomgebieden van Rijn en Rhenum Apud Hermannum Besseling. Maas: modelstudies met Rhineflow-3 en Meuseflow-2; Carthago Consultancy, Snow, C.H., 1917. Wood and other organic structural materials. McGraw-Hill book Rotterdam. company, New York, Londen: Hill publishing co., ltd. Van Deursen, W.P.A., 2006. Rapportage Rhineflow/Meuseflow. Nieuwe Tuente, U., Piepenburg, D., Spindler, M., 2002. Occurrence and settlement of the KNMIscenario’s 2050. Carthago Consultancy, Rotterdam. common shipworm Teredo navalis (Bivalvia: Teredinidae) in Bremerhaven Vrolik, W., Harting, P., Storm Buysing, D.J., van Oordt, J.W.L., von Baumhauer, E.H., harbours, northern Germany. Helgol. Mar. Res. 56, 87–94. 1860.— Verslag over den Paalworm, uitgegeven door de Natuurkundige Van Benthem Jutting, T., 1943. . C. Lamellibranchia.— Fauna van Nederland Afdeeling der Koninklijke Nederlandsche Akademie van Wetenschappen, 12, 1–477. Amsterdam, 153 pp. Van den Hurk, B., Klein Tank, A., Lenderink, G., van Ulden, A., van Oldenborgh, G.J., Katsman, C., van den Brink, H., Keller, F., Bessembinder, J., Burgers, G., Komen G.,